AMD BAREBONE G291-Z20 2U 1XCPU

AMD EPYC™

The New x86 Option

AMD's return to the CPU segment delivers a whole new value proposition to the x86 world. Built on GIGABYTE's extensive design expertise across a range of chipset platforms, this AMD EPYC™ platform is already receiving acclaim for its high specification:

• Memory
• I/O
• Network connectivity

This platform is intended as an alternative for builders in the x86 space and packs a real punch in flexibility and expansion options.

CPU Key Features

Compute
32 cores

Memory
8 channels

I/O
128 PCIe lanes

Power
Regulated per core

Security
Integrated Custom Security

• Memory Optimization
  Memory speed of up to 2666MHz (at 1DPC)
  Total memory capacity of 2TB supported
• PCIe Lanes
  Up to 128 lanes in both 1P and 2P offerings providing the highest expandability options of any x86 platform
• Cores / Power
  CPU power is managed at the core level, meaning that power is allocated only for specific workload needs
• Security
  Proprietary AMD security on chip

Comprehensive I/O

Both single- and dual-socket CPUs come equipped with 128x PCIe lanes, providing extensive I/O options. This is a huge improvement in I/O vs. incumbent x86 offerings.
Increased I/O allows for flexibility in extending your storage, networking or compute to a level previously unattainable with standard server motherboards.
In addition, the integrated nature of the processor - designed as a system-on-chip (SoC) - means that there is no need for a separate PCH and less additional switches needed for expansion lanes.

High Density HPC

Supports 8 x Computing Cards

Able to support up to 8 double slot GPGPU or co-processor cards, the G291 enables world-leading HPC within a 2U chassis. The G291 is ideal for HPC applications such as real time analytics, programs in scientific simulation and modeling, engineering, visualization & rendering, data mining, and much more.

NVIDIA® Tesla® V100 Support

GIGABYTE's AMD EPYC server systems and motherboards are fully compatible and qualified to use with NVIDIA®'s Tesla® V100 GPU, the world's most advanced data center GPU ever built to accelerate AI, HPC, and graphics. Powered by NVIDIA's Volta™ architecture and with 640 Tensor Cores, Tesla® V100 is the world's first GPU to break the 100 teraflops (TFLOPS) barrier of deep learning performance - enabling data scientists, researchers, and engineers to tackle challenges that were once impossible.

High-Speed Networking

GIGABYTE offers 10Gb across dual ports on-board, ensuring that users can get the most out of their system and are well equipped for today's traffic demands. The server's network performance is boosted with full-line rate throughput and CPU overhead is reduced, leaving the EPYC to focus on the real compute tasks.